Transformer-based fast image reconstruction for high fidelity MR knee imaging: an evaluation study

Motivation: Address the challenges of prolonged MRI reconstruction, especially in knee imaging, to enhance clinical efficiency. Goal(s): Develop an efficient knee MRI reconstruction framework utilizing the transformer-based deep learning approach to improve image quality allowing reduced scan time. Approach: Employ the transformer-based framework to reconstruct under-sampled knee MRI data, leveraging transformer-based deep learning for improved image fidelity. Results: Demonstrated enhanced image quality and reduced artifacts compared to conventional methods, showcasing the efficacy in knee imaging reconstruction. Impact: This transformer-based method offers a promising solution for faster MRI reconstruction, potentially improving clinical efficiency and outcomes in musculoskeletal imaging.